| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
Department of Internal Medicine V (U.H.-E., E.C., A.D.), Wilhelminenspital, A-1160 Vienna, Austria; Division of Clinical Endocrinology (G.B.), Hannover Medical School, D-30623 Hannover, Germany; Department of Surgery (H.R.), Sozial-Medizinisches Zentrum Ost, A-1220 Vienna, Austria; Metabolic Unit (G.P.), Institute of Biomedical Engineering, ISIB-CNR, I-35127 Padova, Italy; Ludwig Boltzmann Institute for Leukemia (H.T.), Hanusch Hospital, A-1140 Vienna, Austria; Ludwig Boltzmann Institute for Nutrition and Metabolism (R.P.), Hospital Lainz, A-1130 Vienna, Austria; and 1. Medical Department (M.R.), Hanusch Hospital, A-1140 Vienna, Austria
Address all correspondence and requests for reprints to: Michael Roden, M.D., 1. Medical Department, Hanusch Hospital, A-1140 Vienna, Austria. E-mail: michael.roden{at}akh-wien.ac.at.
 |
Abstract |
|---|
 Top Abstract IntroductionPatients and MethodsResultsDiscussionReferences |
|---|
In conclusion, prolonged weight loss results in a rise of fasting ghrelin concentrations that correlates with fasting insulin concentrations but not improvement of insulin sensitivity.
|
Introduction |
|---|
|
TopAbstractIntroductionPatients and MethodsResultsDiscussionReferences |
|---|
The adipocyte-derived peptide leptin and gut-derived peptide ghrelin exert opposing effects on food intake by decreasing (4, 5) or stimulating (6, 7, 8, 9, 10, 11) appetite and food intake in rodents. Although both hypoleptinemia and leptin resistance are related to body fat mass, the role of leptin in human obesity is uncertain except for rare gene defects (5, 12, 13, 14). Plasma ghrelin (ghrelin) increases before and rapidly falls after meal ingestion, which could result from the rise in plasma glucose and/or insulin concentrations (8, 16). The increase in ghrelin observed before food intake as well as during long-term dieting may therefore represent an adaptation to negative energy balance (10).
In contrast to the poor outcome of long-term dieting in even moderately obese patients (17, 18), several studies demonstrated the long-term efficacy of obesity surgery (17, 19, 20). The mechanisms, however, explaining why surgically treated patients fail to achieve normal body weight are still unclear. Recently it was reported that ghrelin decreased in five obese patients after weight loss by proximal Roux-en-Y gastric bypass (RYGB) surgery (10). It has therefore been hypothesized that a decrease in ghrelin could contribute to the weight loss maintaining effect of this procedure (10), although this remains controversial (21). Measuring ghrelin levels in other bariatric surgical procedures, especially ones that are strictly restrictive and maintain passage of food through the stomach and duodenum, could potentially shed light on ghrelin regulation and help to explain the differences in weight loss efficacy between procedures.
Thus, this study aimed to examine time-dependent effects of prolonged weight loss after a strictly restrictive bariatric procedure on ghrelin and leptin concentrations and their possible correlation with changes in parameters of body mass and insulin sensitivity and secretion derived from the modified oral glucose tolerance test (OGTT) (22) in obesity. We employed laparoscopical adjustable gastric banding (LAGB) (23, 24), which was recently approved by the Food and Drug Administration and allows one to modify the degree of gastric constriction causing no maldigestion but also less dramatic weight loss, compared with gastric bypass surgery (25, 26).
|
Patients and Methods |
|---|
|
TopAbstractIntroductionPatients and MethodsResultsDiscussionReferences |
|---|
Eighteen obese patients (15 females, three males, mean age 35.5 ± 12.0 yr) were included in this study because only patients with an initial body mass index (BMI) greater than 40 kg/m2 were considered eligible for the surgical procedure according to standardized criteria (27). The anthropometric data are summarized in Table 1
.
|
Preoperatively, screening tests were performed to exclude patients with endocrine diseases or general conditions not related to obesity that may reduce longevity. After surgery, regular checks including clinical examination, anthropometric measurements, and routine laboratory tests were performed every 3 months.
In addition to routine checks, endocrine and metabolic tests were performed in identical fashion before, at 6 months, and again 12 months after LAGB. At 1 yr, 14 OGTTs were available because four patients could not be recruited for the third OGTT.
The tests consisted of blood sampling for measurement of fasting concentrations of hormones, peptides, and metabolites along with a modified OGTT for assessment of glucose homeostasis, which was performed in the morning between 0800 and 0900 h after an overnight fast of 8–12 h.
The 2-h plasma glucose value during the OGTT (75 g glucose dissolved in water) was used to define the glucose tolerance status according to the World Health Organization as normal (NGT: 
7.8 mmol/liter or 140 mg/dl), impaired (IGT: > 7.8 mmol/liter), or diabetic (DM: 11.1 mmol/liter or 200 mg/dl). Venous blood samples were collected for insulin, C-peptide, and glucose measurements at fasting and 10, 20, 30, 40, 60, 90, 120, 150, and 180 min after the glucose load.
Thereafter the patients underwent LAGB (Lap-Band, Inamed, Santa Barbara, CA), which benefits from the minimal invasive approach (24). Briefly, a hollow band is connected to a tube attached to a reservoir in the upper abdomen on the anterior rectus muscle sheath. This reservoir permits adjustment of gastric restriction by injection or withdrawal of saline or radiopaque liquid. To warrant nutritional surveillance, patients were seen every 3 months during postoperative follow-up. The previously described nutrition management (24) comprises of a liquid diet for 4 wk after surgery followed by prescription of a solid low-fat diet and a list of rules to avoid vomiting. Nevertheless, energy intake could vary among individuals, depending on their compliance and band calibration. The first inflation of the LAGB is done 6–8 wk after the procedure and further band adjustments were done only in the case of weight stabilization (variation of < 4 kg during the last 2 months) and low vomiting frequency in the outpatient surgical department. Because of individual demand, the frequency of band adjustments were not uniform in patients with LAGB.
The institutional review board approved the study, and written informed consent was obtained from all patients after the nature and potential risks of the protocol had been explained to them.
Assays
Blood was rapidly centrifuged and stored at –72 C until analysis. Routine parameters including lipids and transaminases were measured with standard techniques in the clinical chemistry laboratory. Plasma glucose was measured with the hexokinase method applied on a modular analyzer (Roche, Basel, Switzerland).
Insulin (Serono Diagnostics, Freiburg, Germany), C-peptide (CIS Bio International, Cedex, France), and leptin (human leptin RIA kit, Linco Research, Inc., St. Charles, MO) were quantified with commercial RIA kits with interassay coefficients of variation of less than 5% for insulin and C-peptide and 4.1% for leptin.
Plasma ghrelin was measured by a recently developed RIA (28) whose results closely correlate with the commercially available Phoenix assay (r2 = 0.82, P < 0.0001). Briefly, for generation of ghrelin antiserum, a synthetic C-terminal fragment of ghrelin (amino acids 15–28, 14Tyr, Biotrend, Köln, Germany) was coupled to hemocyanin and administered to rabbits. Employing the iodogen technique, the tracer was purified on a C18 HPLC column and the antibody-bound fraction and free peptide were separated on 2% dextran-charcoal. A solution of full-length ghrelin (Bachem, Heidelberg, Germany) was diluted to yield a standard curve, which strongly correlates with the respective dilutions of ghrelin samples. Ghrelin antibodies exhibit no cross-reactivity with peptides such as leptin, motilin, and GHRH. The detection limit is 34 pmol/liter and the inter- and intraassay coefficients of variation are 4.1 and 2.6%, respectively. This RIA measures total ghrelin including both acylated and des-acyl ghrelin (29).
Calculations
Insulin secretion and appearance. Data obtained during the OGTT were analyzed by a two-compartment mathematical model (22), which, from concomitant analysis of C-peptide and insulin data, allows assessment of 1) basal fasting prehepatic insulin secretion rate per unit volume, BSR, 2) total amount of secreted insulin per unit volume, TIS, and 3) hepatic insulin extraction, HIE (given as percent of the secreted hormone). Details of the model, which was validated in humans against arteriovenous transhepatic measurements (30), are reported elsewhere (22). The areas under the concentration-time curves (AUCs) were calculated with the trapezoidal rule.
Insulin sensitivity.
The OGTT-based index of insulin sensitivity (OGIS) describes glucose clearance during OGTT and is calculated with an explicit formula (31). OGIS has been validated against the glucose clamp and used in other studies (32, 33). The ability to dispose of glucose that relates insulin action to the prevailing systemic insulin concentration is described by the disposition index (34), which is calculated as the product OGIS x 
AUCins, where AUCins is the suprabasal component of AUCins (32). The capacity of the ß-cell to adapt to changes in insulin sensitivity is described by the adaptation index (35), which is calculated as the product OGIS x TIS, where TIS is the suprabasal component of TIS. Both indices provide a quantitative estimate of the ß-cell ability to compensate for insulin resistance.
Statistical analysis
Results are expressed as means ± SD unless stated otherwise. Multivariate analysis of variance and post hoc testing by Dunn-Sidak pairwise comparisons compared differences of mean values. A difference between numeric variables before and after LAGB were analyzed with the paired Student’s t test. In addition, differences in between data on ghrelin were compared with the univariate F test. Pearson’s correlation was used to test the relationship between ghrelin and other parameters. According to the explorative nature of this study, no correction for multiple testing was applied. P < 0.05 was considered to indicate statistically significant differences.
|
Results |
|---|
|
TopAbstractIntroductionPatients and MethodsResultsDiscussionReferences |
|---|
By definition, all subjects were severely obese before LAGB with mean body weights of 127.9 ± 17.1 kg and BMI of 45.3 ± 5.3 kg/m2. They lost a mean of 21.4 kg (–17%) at 6 months and 31.5 kg (–25%) at 12 months after LAGB. Accordingly, BMI dropped by approximately 18 and approximately 26% (both P < 0.0001) and waist circumference fell by approximately 18 and approximately 24% (both P < 0.001) after 6 and 12 months, respectively (Table 1
).
Systolic and diastolic blood pressure decreased (P < 0.0001) by approximately 14% after 6 months and approximately 18% at 12 months after LAGB. Liver enzymes and fasting triglyceride concentrations dropped (P = 0.02), whereas total, low-density lipoprotein, and high-density lipoprotein cholesterol levels were not affected by surgical treatment and weight loss (Table 1).
Metabolites and glucose tolerance status
Fasting blood glucose decreased by approximately 19% at 6 months (P = 0.01) and then remained unchanged until 12 months after LAGB (Table 1).
According to WHO criteria, six patients were diabetic, four had IGT, and eight had NGT before LAGB. After 6 months, three IGT and one DM had converted to NGT, two DM returned back to IGT, and no new case of DM was diagnosed. Of the 14 patients (10 NGT, two IGT, and two DM based on the OGTT at 6 months), who underwent the final OGTT at 12 months, the two diabetic patients DM had returned to IGT, whereas glucose tolerance status remained unchanged in NGT and IGT.
Hormones
Blood sampling for measurements of fasting hormones was taken in the morning between 0800 and 0900 h after an overnight fasting of 8–12 h and immediately before the start of the OGTT.
Fasting plasma insulin decreased from 160 ± 70 to 99 ± 40 pmol/liter at 6 months and to 103 ± 59 pmol/liter at 12 months after LAGB (both P < 0.005 vs. presurgery) (Fig. 1). Simultaneously, fasting plasma C-peptide concentrations decreased continuously by approximately 16 and approximately 32%, compared with values before LAGB (P < 0.008 and P < 0.0001) (Table 1). After 12 months, C-peptide was approximately 19% lower than after 6 months (P = 0.02).
|
).
Plasma ghrelin concentrations were comparable before and after 6 months (234 ± 53 and 232 ± 53 pmol/liter, n.s.) but slightly increased by approximately 13% to 261 ± 72 pmol/liter (P = 0.04 by t test) (Fig. 1). Further time-course analysis revealed a quadratic increase of ghrelin over the complete study period (P = 0.05 after 12 months).
OGTT-derived parameters of insulin sensitivity and secretion
The parameters derived from OGTT analysis are summarized in Table 2. Both indices of insulin secretion, BSR and TIS, decreased during the course of the study with BSR being reduced already at 6 months (P = 0.008). After 12 months, both BSR and TIS were approximately 30% lower, compared with their values before surgery (P < 0.0001 and P = 0.002). HIE increased by approximately 15% after 6 months (P < 0.0001) and then remained unchanged until 12 months.
|
The disposition index was slightly decreased after 12 months, compared with baseline (P < 0.009), whereas the adaptation index did not change during the study (Table 2).
Correlation analyses
Fasting plasma leptin correlated negatively with fasting glucose (r = –0.486, P = 0.04) and hemoglobin A1c (r = –0.506, P = 0.04) and positively with OGIS (r = +0.576, P = 0.01) before LAGB. These relationships disappeared at 6 and 12 months after LAGB.
On the other hand, fasting ghrelin correlated negatively with fasting plasma insulin at 6 (r = –0.495, P = 0.04) and 12 months (r = –0.632, P = 0.02) but not before LAGB (r = –0.295, P = 0.24). Fasting ghrelin correlated positively with HIE at 6 (r = +0.546, P = 0.02) and 12 months (r = +0.571, P = 0.03) but again not before LAGB (r = 0.242, P = 0.34).
Analysis of time-dependent interactions revealed that changes in ghrelin (ghrelin) tended to correlate negatively with changes in body mass (BMI: r = –0.425, P = 0.07) after 6 months and correlated significantly after 12 months LAGB (r = –0.495, P = 0.04). Ghrelin correlated with changes in plasma triglycerides at 6 months (r = –0.567, P = 0.018) but not at 12 months (r = –0.343, n.s.). No significant relationships were observed between fasting ghrelin and other parameters of insulin secretion including fasting C-peptide levels and insulin sensitivity before and after LAGB.
|
Discussion |
|---|
|
TopAbstractIntroductionPatients and MethodsResultsDiscussionReferences |
|---|
Ghrelin, a 28-amino acid peptide, is secreted predominantly by endocrine cells of the stomach and induces weight gain by stimulating food intake in mammals (7, 11, 37). It circulates in an acylated and a desacylated form. Most data are based on measurements of total ghrelin levels. In obese patients, ghrelin concentrations are lower and not suppressed by food intake, compared with normal weight but otherwise matched humans (8, 10). These results suggest that ghrelin is involved in the dysregulation of hunger sensation and long-term body weight in human obesity. In the present study, the obese patients presented with unchanged fasting ghrelin concentrations during the first 6 months after LAGB, although they had lost approximately 18% of their initial BMI. The percent BMI reduction is almost identical with that of approximately 17% induced by a diet protocol of similar duration in patients of similar age but less BMI at study entry (10). Interestingly, the identical percent weight loss by dieting resulted in increased fasting and diurnal ghrelin concentrations similarly measured as total ghrelin. This suggests that gastric restriction by LAGB prevents the early adaptative increase in ghrelin secretion, which would counteract further weight reduction by dietary protocols. A limitation of the present prospective study was the absence of a diet control group. However, recent data suggest that ghrelin is lower after surgical obesity treatment than after dieting (10, 21, 38, 39, 40).
At 12 months, however, fasting ghrelin had increased by approximately 13%, whereas the patients had lost only further approximately 8% of their BMI between 6 and 12 months after LAGB. Because the patients were still obese (BMI approximately 34 kg/m2) at this time, the rise in ghrelin might contribute to the leveling off in weight reduction later on. Such plateau effect and even subsequent weight regain have been reported for LAGB (19). Our observations differ from recent studies on ghrelin concentrations in obese patients who had undergone a proximal Roux-en-Y gastrojejunal anastomosis resulting in bypass of most of the stomach and the duodenum (10, 21). Of note, those studies presented one data set before and after the surgical intervention. Cummings et al. (10) reported that five obese patients after gastric bypass had lower fasting and diurnal ghrelin than normal-weight and matched obese patients. These patients had lost approximately 36% of their initial BMI within a mean observation period of 1.4 yr after surgery, which is similar or slightly higher than the loss of BMI of approximately 26% achieved at 1 yr of the present study. The latter would be in line with observations that gastric bypass surgery is generally more effective to cause a more pronounced and sustained weight loss than gastroplasty (25, 41). In contrast, Faraj et al. (21) reported an approximately 60% increase in fasting ghrelin in obese subjects who already had lost approximately 37% of initial BMI and were still losing weight at approximately 1 yr after the same surgical intervention. On the other hand, these authors observed no change in fasting ghrelin of subjects who were weight stable during the last 6 months of the study. Nevertheless, the present study does not allow to clearly discriminate whether the observed changes are primarily due to either weight loss per se or the surgical technique of LAGB.
We also analyzed the possible contribution of changes in energy balance in our patients. Applying criteria of weight stability of less than 10% weight loss over the last 6 months during follow-up, 11 patients still exhibited a negative energy balance, whereas seven were weight stable. Of note, both subgroups did not differ with regard to fasting ghrelin concentrations (248 ± 55 vs. 282 ± 94 pmol/liter; P = 0.34).
In accordance with our results Holdstock et al. (38) reported higher ghrelin levels after nearly the same weight loss (26 vs. 29%) after 12 months RYGB. In his editorial Cummings (39) summarized that the effect of bariatric surgery on ghrelin levels might depend on the surgical procedure and variable techniques in the surgical departments. In the case of negative energy balance resulting from dieting, the intact gastric fundus, which contains most of the tissues for ghrelin production, might up-regulate ghrelin expression. In line with this contention, Leonetti et al. (40) observed markedly lower postsurgical ghrelin concentrations in RYGB- than in LAGB-treated patients, indicating that the surgical procedures could contribute to the differences in circulating ghrelin. LAGB might differently influence ghrelin because it preserves the stomach in its anatomical site and does not exclude the major ghrelin-producing tissues in the gastric fundus from contact with food.
Interestingly, Cummings et al. (42) reported a 53% increase in ghrelin levels after 8 months LAGB. Although the study period was shorter than in our study, the reported increase in ghrelin concentration was significantly higher that the 13% we observed after 1 yr LAGB. One explanation could be the different pouch size because we tried to reduce the pouch to a minimum to avoid pouch dilatation. Alternatively, the activity of the (para)sympathic nerve system could be involved in ghrelin secretion and/or action. The contribution of the vagal nerve activity to secretion of ghrelin has been recently studied in experimental animals. Date et al. (43) confirmed the presence of ghrelin receptors in the vagal nerve and transport of ghrelin receptors through vagal neurons to the periphery. Although vagotomy resulted in reduction of food intake, the meal-dependent response of ghrelin was not affected in rats (44). In humans, Nissen fundoplication, a surgical technique that results in a reflux barrier, does not affect vagal nerve function, although relaxation of the lower esophagus and fundus relaxation was impaired (45). Reduced relaxation of esophagus was also demonstrated in patients treated with LAGB (46), causing increased satiety and a decrease in hunger, appetite (47), and prolonged reduction of caloric intake (48). Nevertheless, it is not known to what extent LAGB directly affects vagal nerve signaling in humans.
The present study also found a strong negative correlation between fasting ghrelin and insulin after LAGB. This is in line with accumulating evidence that insulin is a significant independent determinant of fasting ghrelin concentrations (49). Although the mechanisms by which insulin influences ghrelin release are yet unclear, components of the insulin signaling cascade were identified in the gastrointestinal tract (50, 51). Thus, insulin could decrease ghrelin by directly inhibiting ghrelin release from stomach cells. In this study, fasting plasma insulin and C-peptide concentrations and parameters of insulin secretion were reduced after LAGB that could have therefore lead to des inhibition of gastrointestinal ghrelin secretion. Nevertheless, the lack of a correlation between ghrelin and fasting plasma C-peptide as well as model-derived parameters of insulin secretion indicates that reduction in insulin secretion is unlikely to be responsible for ghrelin secretion after LAGB. On the other hand, ghrelin concentrations were positively associated with HIE, which reflects insulin degradation in the liver and will also result in lower circulating plasma insulin concentrations. Accelerated hepatic insulin degradation after LAGB could be a byproduct of weight loss, which ameliorates nonalcoholic fatty liver disease in obesity (52). Of note, liver function improved as indicated by reduction in liver enzymes such as glutamic pyruvic transaminase and 
GT at 12 months after LAGB (53). Finally, increased ghrelin could decrease plasma insulin concentrations in obese patients as indicated by reduction of plasma insulin upon ghrelin administration (54).
Nevertheless, indirect effects mediated by the central nervous system cannot be excluded because both ghrelin and leptin cross the blood-brain barrier (55). Ghrelin induces hunger by increasing the expression of neuropeptide Y, which is inhibited in turn by insulin and leptin (56). The present study confirmed that circulating leptin concentrations are increased in obesity and decreased by weight loss (57, 58), which is mostly but not entirely accounted for by reduced body fat mass (59). Fasting plasma leptin correlated positively with indicators of insulin sensitivity such as OGIS and negatively with fasting plasma insulin before LAGB. This is in line with reported insulin-like effects of leptin on peripheral metabolism (60, 61, 62). Interestingly, we found no correlation between leptin and either BMI or waist circumference before surgery. This is partly in line with a recent study performed in a population of similar degree of obesity, which reported no correlation between leptin and BMI but a positive correlation between leptin and fat mass (63). We also found no correlation between leptin and either BMI or waist circumference at 6 and 12 m after LAGB. This is different from recent data showing a positive correlation between leptin and BMI at 6 (63) and 15 months (21) after RYGB.
In conclusion, prolonged weight loss results in a rise of fasting ghrelin concentrations, which correlate negatively with fasting insulin concentrations but not with insulin sensitivity. The decrease in fasting insulin concentrations could be due to increased hepatic insulin extraction rather than altered insulin secretion.
|
Footnotes |
|---|
Abbreviations: AUC, Area under the concentration-time curve; BMI, body mass index; BSR, basal fasting prehepatic insulin secretion rate per unit volume; DM, diabetes mellitus; ghrelin, changes in plasma ghrelin; HIE, hepatic insulin extraction; IGT, impaired glucose tolerance; LAGB, laparoscopical adjustable gastric banding; NGT, normal glucose tolerance; OGIS, OGTT-based index of insulin sensitivity; OGTT, oral glucose tolerance test; RYGB, Roux-en-Y gastric bypass; TIS, total amount of secreted insulin per unit volume.
Received August 18, 2003.
Accepted March 29, 2004.
|
References |
|---|
|
TopAbstractIntroductionPatients and MethodsResultsDiscussionReferences |
|---|
system in morbidly obese women after gastric bypass. Obes Surg 13:615–621[CrossRef][Medline]This article has been cited by other articles:
 |
|
 |
|
  E. Ferrannini and G. Mingrone Impact of Different Bariatric Surgical Procedures on Insulin Action and {beta}-Cell Function in Type 2 Diabetes Diabetes Care, March 1, 2009; 32(3): 514 - 520. [Full Text] [PDF]
|
|
|
|
 |
|
 N. Garcia de la Torre, M. A. Rubio, E. Bordiu, L. Cabrerizo, E. Aparicio, C. Hernandez, A. Sanchez-Pernaute, L. Diez-Valladares, A. J. Torres, M. Puente, et al. Effects of Weight Loss after Bariatric Surgery for Morbid Obesity on Vascular Endothelial Growth Factor-A, Adipocytokines, and Insulin J. Clin. Endocrinol. Metab., November 1, 2008; 93(11): 4276 - 4281. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 A. Arepally, B. P. Barnett, T. T. Patel, V. Howland, R. C. Boston, D. L. Kraitchman, and A. A. Malayeri Catheter-directed Gastric Artery Chemical Embolization Suppresses Systemic Ghrelin Levels in Porcine Model Radiology, October 1, 2008; 249(1): 127 - 133. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
M. Shah, V. Simha, and A. Garg Long-Term Impact of Bariatric Surgery on Body Weight, Comorbidities, and Nutritional Status J. Clin. Endocrinol. Metab., November 1, 2006; 91(11): 4223 - 4231. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 D. L. Williams and D. E. Cummings Regulation of Ghrelin in Physiologic and Pathophysiologic States J. Nutr., May 1, 2005; 135(5): 1320 - 1325. [Abstract] [Full Text] [PDF]
|
|
| |||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
| Endocrinology | Endocrine Reviews | J. Clin. End. & Metab. |
| Molecular Endocrinology | Recent Prog. Horm. Res. | All Endocrine Journals |
